Determination of spatiotemporal resolution in one-way nesting of ocean regional circulation model

Dynamic downscaling approach has been developed to nest relatively coarse information from the Ocean Global Circulation Models (OGCMs) to operationally forecast or project future climate information with high resolution in the regional scales. Nesting procedure using the Ocean Regional Circulation Models (ORCMs) to OGCMs degrades the quality of predictions due to many sources of error, in particular small scale motions. This study evaluates the impacts of errors due to the grid differences, updating times, and domain sizes with the Big-Brother Experiment (BBE) to recommend the optimal nesting strategy to minimize errors. The BBE test has advantage in evaluating each error separately depending its source and so the qualities of the ORCMs’ reproduction are evaluated for each error sources. In additions, the results of the nestings under the various conditions are analyzed with Taylor diagram to deduct optimal conditions for the grid size jump ratios, the updating frequencies and the domain sizes. The optimal ratio of spatial resolutions between original data and nested model ones are obtained at 3 or less spatial resolution and the information of the Lateral Boundary Conditions (LBCs) should be updated at every 6 h considering computational cost and storage. Since errors produced at the LBCs propagate into the center of the nested model domain, extra space is required to reduce the effect of the error to the domain of interest. However, the determination of extra space between LBCs and interest area is related to computational cost. Thus if the ORCMs’ domain size is larger than twice of the domain of interest, then the errors from the LBC can not reach the domain of interest.